With the advent of the sound sources compliant with multichannel audio reproduction, such as DVD (Digital Versatile Disc), BS digital broadcasting, etc., in order to allow a user who has only a 2-channel loudspeaker system to enjoy multichannel reproduction, various audio processors which realizes multichannel reproduction in a virtual manner only with front loudspeakers have been developed. Specifically, in such an audio processor, to reproduce audio signals for rear channels through loudspeakers placed in front of a listener, the audio signals are subjected to sound image localization such that the sound image position which is perceived by the listener is at the rear of the listener (virtual surround processing).
A known example of such an audio processor is an audio processor 4000 which adds together signals obtained by performing the virtual surround processing on rear-channel audio signals (virtual surround signals) and front-channel audio signals to output audio PCM signals for the front loudspeakers as shown in FIG. 12 (see Patent Document 1 and Patent Document 2).
The audio processor 4000 includes a virtual surround processing section 4100, a volume normalizing section 4200, and an adder 4300.
Among externally-input audio PCM signals of 4 channels in total, i.e., front-channel audio signals for front left and right reproduction (2 channels for left and right) and rear-channel audio signals for rear left and right reproduction (2 channels for left and right), the virtual surround processing section 4100 performs the virtual surround processing on the rear-channel audio signals to output virtual surround signals.
The volume normalizing section 4200 performs a process on the front-channel audio signals and the virtual surround signals such that the volume levels of the signals are within a predetermine level range (volume normalization). The volume normalization is performed for the purpose of preventing occurrence of an overflow in the addition of the front-channel audio signals and the virtual surround signals.
The adder 4300 adds together the front-channel audio signals and the virtual surround signals which have been volume-normalized by the volume normalizing section 4200.
When front-channel audio signals and rear-channel audio signals are input from the outside of the thus-constructed audio processor 4000, the virtual surround processing section 4100 performs the virtual surround processing on the front-channel audio signals to output virtual surround signals to the volume normalizing section 4200. After the volume normalizing section 4200 performs the volume normalization on the front-channel audio signals and the virtual surround signals, these normalized signals are added together by the adder 4300 and output as audio PCM signals for front loudspeakers.
As described above, in the audio processor 4000, the virtual surround processing is performed on rear-channel audio signals, whereby multichannel reproduction is realized in a virtual manner only with front loudspeakers.
Another example of an audio processor which realizes multichannel reproduction in a virtual manner is an audio processor 5000 wherein reflection sound creation processing which creates reflection sound in a virtual manner is performed on 2-channel audio PCM signals for front left and right reproduction (2-channel stereo audio signals) which are input from the outside of the audio processor to generate pseudo rear-channel audio signals (see, for example, Patent Document 3).
The audio processor 5000 includes a virtual surround processing section 4100, a volume normalizing section 4200, an adder 4300, and a reflection sound processing section 5400.
The reflection sound processing section 5400 performs a process which creates reflection sound in a virtual manner (reflection sound creation processing) on 2-channel stereo audio signals to output rear-channel audio signals.
In the thus-constructed audio processor 5000, the reflection sound processing section 5400 performs the reflection sound creation processing on 2-channel stereo audio signals to output pseudo rear-channel audio signals. These rear-channel audio signals are subjected to the virtual surround processing in the virtual surround processing section 4100. The virtual surround-processed signals and the stereo audio signals of the channels are subjected to volume normalization in the volume normalizing section 4200 and added together at the adder 4300.
That is, in the audio processor 5000, 2-channel stereo audio signals are subjected to sound image expansion processing and added to the original signals, whereby multichannel reproduction that achieves a stereophonic effect in a virtual manner is realized.                [Patent Document 1] Japanese Laid-Open Patent Publication No. 6-233394        [Patent Document 2] Japanese Laid-Open Patent Publication No. 10-174197        [Patent Document 3] Japanese Laid-Open Patent Publication No. 8-116597        